7867 Treatment-Related Changes in Total Hip Bone Mineral Density and Fracture Risk Reduction for Drugs With Different Mechanisms of Action: The FNIH-ASBMR-SABRE Project

Abstract Disclosure: T. Vilaca: None. M.T. Schini: None. A.R. Thompson: None. E. Vittinghoff: None. L. Lui: None. S.K. Ewing: None. D.C. Bauer: None. D.M. Black: None. M. Bouxsein: None. R. Eastell: Consulting Fee; Self; Immunodiagnostic Systems, CL Bio, CureTeQ, Biocon, Takeda, UCB. Grant Recipient; Self; Alexion Pharmaceuticals, Inc., Osteolabs. Speaker; Self; Pharmacosmos, Alexion Pharmaceuticals, Inc., Amgen Inc, UCB. Introduction: To date, trials of new anti-osteoporosis therapies have used fractures as primary endpoints. In the FNIH-ASBMR-SABRE project, we posited that the treatment-related change in total hip BMD (THBMD) could be used as a surrogate endpoint for fractures in future clinical trials of new osteoporosis therapies. To demonstrate this, we compiled a large dataset of individual patient data (IPD) from randomized trials and showed a strong association between the change in THBMD at 24 months and reduction in fracture risk. To further explore the use of THBMD as a surrogate endpoint, here we examine whether the association is robust across drugs with different mechanisms of action. Methods: We used IDP (n>120,000 participants) from 22 randomized, placebo-controlled, double-blind trials of osteoporosis medications (17 anti-resorptive, 3 PTH analogs, 1 odanacatib, and 1 romosozumab). We calculated the treatment-related difference (active-placebo) in mean % change in THBMD at 12, 18, and 24 months for each trial. We determined the treatment-related fracture risk reductions for the entire follow-up period, using logistic regression for radiologic vertebral fractures and Cox regression for all clinical fractures (a combination of non-vertebral and clinical vertebral fractures). We used meta-regression to estimate the study-level association (r2) between treatment-related differences in THBMD changes and fracture risk reduction, including only the 17 trials of anti-resorptive drugs and then using the entire set of 22 trials. Since there were only 3 trials of anabolic drugs, we did not perform separate analyses for the anabolic trials. Results: The r2 values were similar between all trials and antiresorptive trials only for both fracture outcomes at all time points. Specifically, for vertebral fractures, the r2 was 0.59 vs 0.70 at 12 mo, 0.69 vs 0.74 at 18 mo, and 0.73 vs 0.78 at 24 mo for all trials vs anti-resorptive trials only. For all clinical fractures, the r2 was 0.46 vs 0.51 at 12 mo, 0.64 vs 0.60 at 18 mo, and 0.71 vs 0.65 at 24 mo for all trials vs anti-resorptive trials only. Conclusion: These results demonstrate that the associations between treatment-related changes in THBMD and fracture risk reductions are robust across anti-osteoporosis therapies with varying mechanisms of action. We conclude that the treatment-related difference in THBMD change would predict fracture reduction equally well for drugs with different mechanisms of action. Presentation: 6/2/2024